1. Field of the Invention
The invention relates to a technology that controls controlling subjects.
2. Description of the Background Art
Conventionally, controllers that control controlling subjects have been controlled to implement most suitable control without any effects, e.g. of aged deterioration and of individual differences of the controlling subjects. For example, in the controllers, control values (learning values), taken into account differences of the past between target values of the control and actual measurement values that have actually controlled the controlling subject based on the target value of the control (offset values), are written in non-volatile memory in advance. On the next control, the learning values of the non-volatile memory are loaded into volatile memory, where the controlling subjects are controlled based on learning values of the volatile memory, what is called, learning control is applied.
When controllers implement control, the volatile memory with faster access speed is more suitable than the non-volatile memory as a storage media for deleting and writing of parameters including leaning values. However, the volatile memory has a defect that stored data are lost when electric power is suspended. Therefore, a specific volatile memory (hereinafter referred to as specific memory) is used to implement a speedy control, to which the power is constantly supplied from the power supply even when electric power to the controllers is suspended.
Even such a specific memory in the controller has risks like data are deleted when a line for electric power is taken off at repair or when electric supply is suspended at accidents. Further, there is another risk that electronic noise might cause data abnormality. As a result, the method, writing parameters written in the specific memory into the non-volatile memory as backup data, is applied.
As such a technique, it is known that the backup data at the non-volatile memory are reloaded into the specific memory when abnormalities occur at the specific memory in the controller for example.
However, even the stored data in the non-volatile memory have risks that data abnormalities occur when the non-volatile memory breaks down. As a result, an appropriate control is not implemented if the controlling subjects are controlled by using the abnormal data.